Generative AI as a Digital Learning Companion: Opportunities and Risks in Higher Education

Abstract

This paper examines the dual-edged role of generative artificial intelligence as a digital learning companion in higher education, critically analysing its transformative opportunities alongside significant ethical and pedagogical risks. The primary objective is to synthesise current research to understand how GenAI tools—such as ChatGPT, Gemini, and AI-integrated learning platforms—can enhance personalised learning, academic writing, accessibility, and student engagement, while simultaneously identifying and addressing concerns related to academic integrity, cognitive overreliance, algorithmic bias, misinformation, data privacy, and equitable access. The study employs a comprehensive literature review methodology, drawing on recent empirical and conceptual publications from 2023 to 2025. Key findings reveal that GenAI significantly improves learning efficiency, motivation, and inclusivity—particularly for non-native speakers, neurodiverse learners, and those in under-resourced settings—yet its benefits are contingent upon human-in-the-loop design, teacher supervision, and theory-informed implementation. The study concludes that GenAI’s greatest value lies not as a replacement for educators but as an augmentative, ethically governed partner within human-centred pedagogical frameworks. These insights underscore the urgent need for coordinated policy development, AI literacy training, and collaborative governance to responsibly harness the potential of GenAI in advancing equitable, innovative, and pedagogically sound higher education.